Mixing in Ableton Live 12 with Stock Devices: A Small Worked Mix

A kick can sound huge alone and disappear when the bass enters. A bright piano can feel clear in solo and cover the lead in the chorus. Mixing in Ableton Live starts with deciding which sound deserves your attention, then making the others support it.

We’ll mix a small instrumental in Live 12.4.15 Suite using stock devices and Core Library sounds. The exact presets and device chain assume Suite. Other editions have different device and instrument libraries; check which tools you own before rebuilding the example. See Ableton’s edition comparison. Settings are starting points to audition. If the interface is new, keep our Ableton beginner’s guide beside you.

1. Build a mix you can actually judge

Use six MIDI tracks: Kick, Snare, Hat, Bass, Keys and Lead. Load Kick 909 1, Snare 909 Hard 1 and Hihat Closed 909 into separate Simplers; trigger each at MIDI 60 so it plays at its original pitch. Load Basic Analog Bass, E-Piano Classic and Basic OG Lead on the remaining tracks. Separate drums make level and sidechain controls easy to reach.

Set 120 BPM and loop four bars in 4/4. Kick plays every quarter; snare plays beats 2 and 4; hat plays eighths. Our Drum Patterns: Sixteen Steps, Four Beats connects separate drum layers to a counted grid. Its printed hat strikes are offbeats only; this mix adds a hat on each beat too. The hip-hop guide shows how changing attacks changes the groove before processing enters the picture.

The fixed harmonic material for the worked mix: one chord per bar and a separate root below it. The bass notes are A1, F2, D2 and E2 in scientific pitch names; their lengths are a dotted quarter and a dotted eighth, with rests completing each bar. These lower pitches are intentional: octave choice gives the bass a role before any EQ is added.

Use Am–F–Dm–Em, one chord per bar. The keys play A3–C4–E4, A3–C4–F4, A3–D4–F4, G3–B3–E4 in scientific pitch names. Bass plays A1, F2, D2, E2: each root starts on beat 1 for 1½ beats, then beat 3 for ¾ beat. Live labels MIDI 60 C3; the notation calls it C4. Voice a Chord Loop in MIDI explains the chord inversions and separate bass lane.

The bass clip in Live 12.4.15. Each root starts on beat 1 for 1.5 beats and beat 3 for 0.75 beat. Live’s A0/F1/D1/E1 labels are one octave below the scientific names. Note-editor crop includes velocities; mixer and clip controls are omitted. Preset release can ring into the gaps.

Give the lead two short notes per bar, on the “&” of 1 and 3: E5–C5, F5–C5, F5–D5, E5–B4. For this version, use velocity 100 for kick, snare and lead, 85 for keys, and 100 then 90 for the two bass attacks in each bar. Alternate hat velocities 95 on the beat and 80 on the offbeat. Kick and snare MIDI notes last 0.2 beat; hats last 0.15 beat; lead notes last half a beat. Drum note lengths are MIDI gates—the sample’s playback mode and envelope determine how long you hear it. Hold each keys chord for the whole bar.

The lead answers between the quarter-note kicks: its two attacks fall on the “&” of 1 and the “&” of 3 in every bar. Read the rests as part of the arrangement. They let you hear the delay repeats and the room tail separately from the next direct note, which makes it easier to judge whether a send is helping or blurring the phrase.

Choose a reference track whose lead and drum relationship you like. When the reference plays through Main, bypass the processing you added to Main for that comparison. Solo the reference and the mix alternately; never play them together. Match their listening levels with the reference track’s gain, then restore the mix’s Main chain when you return to your work.

Save a dry version. Let the lead carry the tune and the kick mark the pulse.

The order of decisions in this article. First establish the tune and pulse with levels and pan; then use EQ and dynamics for a remaining problem; then add shared space and judge the combined sound. The last comparison uses matched listening level so that extra gain does not get mistaken for a clearer balance. You can stop at any stage that already sounds convincing.

2. Gain staging: leave room and keep comparisons fair

Gain staging means managing level along the signal path. Berklee’s overview explains why the next processor needs a suitable input. Put Utility immediately after each instrument and begin with Gain at −6 dB. This is convenient trim for this exercise, not a universal calibration level for Live’s instruments.

A simplified track signal path. Utility comes after the instrument but before the mix processors in this example; the track fader follows the device chain. If the next processor responds too strongly, change its input or its own settings. If the instrument only dominates the musical balance, change its track fader. These actions solve different problems even when both make something quieter.

The source trim sets what enters later devices; the track fader sets the part’s balance after them. Lowering a fader cannot undo distortion already created upstream. For recorded audio, clipping at the interface also survives a quieter playback level; Focusrite’s gain guidance explains leaving room for unexpectedly loud playing.

Source trim in Live 12.4.15: E-Piano Classic → Utility → EQ Eight. The outlined Utility shows Gain −6.00 dB. The EQ's 300 Hz cut shown here is the later §4 move. The mixer and unused device area are cropped away, preserving device order. This is the exercise trim, not a universal instrument-level requirement.

Play the busiest passage with Main at 0 dB and its processing bypassed. Lower tracks until Main peaks comfortably below zero; roughly −6 dBFS is a working aim here. Headroom is the distance to the digital ceiling, not a compulsory mastering handoff value.

Live’s Audio Fact Sheet describes floating-point mixing headroom: a red internal meter does not automatically prove permanently clipped audio. Keep levels manageable because processors can respond differently to input level, and the hardware output has limits.

For one jumping note, try velocity editing before compression. For a gradual section lift, use track-volume automation. Fix a performance or section directly when possible.

When bypassing processing, match its output loudness to the bypassed version. Keep monitoring comfortable and consistent so extra gain does not bias the comparison.

3. Balance and pan before carving frequencies

Pull the faders down, then bring up kick and bass together. Check whether you can hear both the kick’s strike and the bass’s changing pitches. Add snare, then hat quietly enough that it supports the pulse without becoming the lead. Add the melody; bring keys up only until their harmony becomes clear.

Our initial fader positions are Kick −10, Snare −14, Hat −22, Bass −14, Keys −20 and Lead −18 dB, after the Utility trims. These are fader settings, not measured output peaks. Envelopes and velocities affect the result; rebalance by ear and check Main.

The mixer in Live 12.4.15 before grouping. Tags 1–3 mark tracks, returns and sends; tag 4 marks both Post switches; tag 5 marks the return faders at 0 dB. Empty middle columns and the tall empty clip area were cut out. The Main fader position in this capture is not part of the balance exercise; keep Main at 0 dB when following the method. Playback is stopped; the meters do not establish playing levels.

Keep kick, bass, snare and lead centered. Try keys 15% left and hat 20% right. Use normal stereo pan here; a preset’s existing width still matters. Live’s Mixing chapter explains the pan modes.

A proportional pan plan: Keys is 15% of the center-to-left distance and Hat is 20% of the center-to-right distance. Both moves stay near center; the larger right move is drawn farther out. Existing stereo width still depends on the presets. Check the result in stereo and mono.

Judge balance with the whole loop playing. Solo can locate a ring or click, but removes the competition that prompted the change.

Turn the monitoring down briefly. If the tune vanishes, lower the supporting keys before boosting the lead’s treble. Place Utility on Main and enable Mono to combine left and right; Bass Mono checks only the low range. If a part loses much more presence than the others, solo it briefly with Mono still on, then bypass its widening, chorus or stereo-delay stages one at a time. Restore the full mix to judge the fix, then turn Mono off for the final stereo listen.

If two parts compete every time they play together, change an octave, shorten a note or remove a layer. The Build a Bassline from Chord Roots exercise makes you compare bass attacks with kick attacks. Arrangement can open space that EQ cannot carve gracefully.

4. EQ Eight: solve one overlap at a time

Now ask whether the keys’ low body still hides the bass. On Keys, add EQ Eight after Utility. In Stereo mode, turn off unused bands, enable band 3 and choose its bell-shaped filter. Select 3, then enter 300 Hz, −3 dB and Q 1 in the left controls; turn Adaptive Q off to match the picture. Frequency locates the cut, Gain sets its depth and Q sets its width: higher Q makes this bell narrower. Listen in the full mix: does the bass pitch become easier to follow while the keys retain enough body? Keep it only if separation improves and the lead stays clear; otherwise undo it.

EQ Eight in Live 12.4.15: selected bell band 3, 300 Hz, −3 dB, Q 1; Adaptive Q off. The outline marks the band’s controls. Device-only crop. Keep the cut only if it improves separation at matched level.

300 Hz is a search point for these parts, not a universal “mud” frequency. EQ Eight’s reference explains the filters; judge their usefulness in the mix.

Match the width of an adjustment to the audible problem. A broad excess of body may respond to a broad bell; an isolated ringing pitch may call for a narrower treatment. These equal-depth curves are explanatory drawings, not screenshots or response measurements, and they do not prescribe two simultaneous cuts. The worked EQ Eight setting uses only the broad −3 dB proposal.

Use a broad cut for a broad excess of body. Reserve a narrow cut for a specific ringing pitch you can actually hear. A huge boosted sweep can make almost any frequency annoying; keep the audition level low and return to modest settings before judging the change.

If you hear unwanted low energy in the keys, audition EQ Eight’s 12 dB/octave low-cut filter around 100 Hz. Raise its frequency cautiously, then back down if the chords lose useful body. Do not apply that same cutoff to the bass.

Use the analyzer to locate an audible concern, not to pursue a target shape. Hide it briefly to judge the adjustment by ear.

Leave kick and bass EQ flat initially. If both demand the same low-frequency space, first decide which should carry the deepest weight. Cut the supporting part only where that improves the combination. Bypass in context, compensate level, and stop when the line becomes readable; eight available bands do not require eight decisions.

5. Compressor and Glue: control movement without flattening it

Name the dynamics problem first: jumping peaks, lost attack or disconnected drums. For a consistently quiet lead, raise its fader first.

A compressor changes gain according to a detector signal. Threshold sets the level around which compression begins. At 3:1, a steady detector level 3 dB above threshold produces roughly 1 dB of output increase above it, once the compressor has settled. The pictured soft knee makes the onset gradual. Attack controls how quickly reduction builds; release controls how quickly it relaxes as the detector level falls. The Compressor reference explains the controls. For our lead, start at 3:1, attack 20 ms, release 100 ms, full wet, with automatic makeup off. The lead screenshot uses RMS detection and an enabled high-pass detector filter at 80 Hz, resonance 0.71. That filter changes what triggers compression, not the lead’s audio tone directly. Leave it off initially unless low energy is driving unwanted reduction; enable it to reproduce the pictured setup. Lower threshold until louder notes show about 2–3 dB reduction.

Lead Compressor in Live 12.4.15. Tag 1 marks ratio 3:1, attack 20 ms and release 100 ms; tag 2 marks the −24 dB starting threshold. Makeup is off and Dry/Wet is 100%. Device-only crop; playback is stopped. During playback, lower threshold only if 2–3 dB of reduction improves the phrase.

The prepared −24 dB threshold must follow the actual input. If notes are already even, bypassing may be better. Match output level before comparing.

Separate two listening questions: what happened to the leading edge, and how does the level recover afterward? The curves illustrate one possible outcome rather than a measured response to the article’s lead. Attack is not simply a period during which a compressor does nothing, and release is not a fixed silent gap. The source envelope, detector and settings shape what you actually hear.

A slower attack can preserve more of the leading edge; a faster one can soften it. Neither setting guarantees punch. If the melody feels dull, lengthen attack before boosting highs. If it stays pinned down between notes, shorten release or reduce the amount of compression. Listen for the phrase’s movement, not just a tidy reduction meter.

Compressor offers flexible response controls; Glue’s stepped timing and ratios suit auditioning shared drum movement here. Both work on individual parts or buses. Compare their response at matched level.

For drums, select Kick, Snare and Hat and use Edit’s Group Tracks command. Name the group Drums. Grouping normally sends those tracks into the group, where one fader and one device chain affect their combined sound. Confirm their outputs point to Drums, especially in an existing Set with custom routing.

Drums in Live 12.4.15. Tags mark child outputs feeding Drums and Glue’s 2:1/10 ms/Auto controls. Threshold −12 dB and Makeup 0 dB are starting values; playback is stopped. Other tracks/returns are cropped away; an empty vertical strip between headers and mixer is removed.

Try Glue Compressor on Drums: ratio 2:1, attack 10 ms, release Auto, Dry/Wet 100%, Soft Clip off. Lower threshold for only 1–2 dB reduction on stronger hits, then match with Makeup. Shared gain movement may help the drums feel connected; if the hat sinks on every kick or the snare loses its front, reduce compression or bypass it. “Glue” names a device, not a guaranteed improvement.

6. Sidechain the bass only as far as needed

Sidechain compression lets the kick control gain reduction on the bass. Put Compressor on Bass after Utility. On Bass, unfold Compressor’s sidechain panel with its disclosure button. Enable the external sidechain, choose Kick in the source chooser and Post FX in the tap chooser. Keep sidechain listening off for normal playback. Post FX follows the kick’s devices without following its track fader. The Routing chapter explains these routing tap points.

Bass Compressor in Live 12.4.15. Tags 1–3 mark External on, Source Kick and Post FX. Mix is 100% and SC Gain is 0 dB; response starts at 4:1, 2 ms attack, 120 ms release and −24 dB threshold. Device-only crop; playback is stopped, so this shows routing and settings rather than achieved reduction.

Begin at 4:1, attack 2 ms, release 120 ms, automatic makeup off and full wet. During playback, try lowering threshold for about 2–4 dB reduction at kick attacks. Check whether that gives the strike space while the bass notes remain clear. The detector listens to the kick; the sound being turned down remains the bass.

Ducking makes a temporary opening in time. This drawing shows gain movement rather than the bass waveform and does not specify an exact amount or duration of attenuation. There are 500 ms between quarter-note kicks at 120 BPM, but the 120 ms release control does not define the entire recovery curve. Judge whether the bass returns naturally before the next kick.

At 120 BPM the kicks are 500 ms apart. That gives you a useful listening reference, but 120 ms release does not mean exactly 120 ms of silence. Shorten release if the bass fails to recover between hits; lengthen it if the low notes chatter or the level snaps back unnaturally. Reduce the depth if the bass seems to vanish.

Both bass attacks—on beats 1 and 3—coincide with kicks. Try moving the beat-3 attack between kicks before increasing compression, keeping its duration and the bar length intact. Check all four roots: one threshold can produce an uneven musical result across different notes and envelopes.

Disable sidechain listening before judging the mix. Compare with the compressor bypassed at a similar perceived bass level. Keep the ducking only if you can follow both kick and bass more easily. Strong audible pumping is a creative option, but it is a different goal from this mix’s small opening.

7. Sends and returns: share a room, leave space for echoes

Use Create > Insert Return Track until you have two returns. Name them A Room and B Echo, load Reverb on A and Delay on B, and route both to Main. Start both return faders at 0 dB and all track sends down. Set each return’s Pre/Post toggle to Post in the mixer’s Main section; those switches govern the sends feeding that return. Set both effects fully wet. A send adds a parallel effect path; the track retains its direct sound.

Why the return device is fully wet in this example: the direct lead already reaches Main through its own track, while Return B adds echoes on a separate path. The send chooses that track’s contribution to the delay. The return fader sets the combined effect output, which matters when several tracks share it. Return A follows the same parallel principle with Reverb instead.

Start Reverb with 0.8-second decay and 15 ms predelay. Send a little snare, keys and lead into it; leave kick and bass dry initially. Predelay gives the direct attack a small lead over the reverberation. If the mix feels distant, reduce the sends first; if successive chords smear together, try a shorter decay.

Return A’s Reverb in Live 12.4.15. Tags mark 15 ms predelay, 800 ms decay and Dry/Wet 100%. Device-only crop; the mixer is omitted. Track sends set how much enters this shared room.

Mute the room return, then restore it. Keep it only if the shared space helps; an available send need not be used.

For Delay, try 250 ms on both channels and 20% feedback. At 120 BPM that is an eighth-note interval. The Tempo Math: Beats, Bars and Delay Times worksheet derives it from 60,000 divided by BPM, then halved. Time mode states milliseconds; a synced division follows future tempo changes. This example keeps the tempo fixed.

Return B’s Delay in Live 12.4.15. Tags mark Time mode at 250 ms on both channels, Feedback 20% and Dry/Wet 100%. Device-only crop; mixer omitted. At 120 BPM this is eighth-note spacing, but fixed milliseconds will not follow a tempo change.

Raise only the lead’s B send until you notice an answer in its gaps, then back it down slightly. If the next note is obscured, lower feedback or send level before reaching for EQ. Lowering the return fader changes the entire shared effect; lowering one track’s send changes only that contribution. Keep return-to-return sends off for this exercise so the routing stays easy to hear.

8. A light Main chain and a useful final check

Start with Utility for the temporary mono check, then Limiter. Keep Utility neutral for the final stereo listen. Leave EQ and compression off Main unless you can identify a mix-wide problem. A boomy bass part calls for returning to Bass, rather than thinning every instrument with a Main EQ.

Main in Live 12.4.15: neutral Utility → Limiter, Input Gain 0 dB, Maximize off, Ceiling −1 dB, True Peak, 3 ms lookahead and Auto release. Device-only crop; mixer omitted and playback stopped. These settings do not establish loudness or achieved limiting.

On Limiter, leave Maximize off, choose True Peak, set Ceiling to −1 dB, use 3 ms Lookahead and Auto release. Start Input Gain at zero; increase it only if you need a louder listening copy. Aim for occasional 1–2 dB reduction, not continuous heavy limiting. This ceiling is a conservative example, not a universal delivery specification. ITU-R BS.1770 treats loudness and true peak as separate measurements; a peak ceiling does not establish a loudness target.

Three checks after the four-bar exercise. Bypass with level matching tests whether processing helped. Mono tests whether the musical anchors remain understandable when the stereo channels combine. Listening through the full song tests changes and tails that a short loop can conceal. If one check fails, return to the relevant part before adding another processor to Main.

Keep Main at unity when checking Limiter’s ceiling: post-device level changes affect the output. Adjust monitoring for a fair comparison.

If the snare flattens or bass distorts, back off the gain and revisit the mix. Compare against your dry version at matched listening level, then check the full arrangement, quieter sections, transitions and effect tails. A four-bar loop teaches decisions; it cannot prove a whole song is finished. Finish with two clearly labeled files:

1. If you built Session clips, copy all six into Arrangement at bar 1. In Arrangement View, select the song and enough extra time for its final reverb and delay to fade.
2. Ensure Arrangement playback is active, then choose File > Export Audio/Video and render Main. Leave Normalize, Convert to Mono and Render as Loop off for these whole-song copies.
3. Make one stereo WAV with the listening Limiter enabled and another with it bypassed for handoff. Listen to the exported files, including their endings, and label the two versions clearly.

The export manual explains the dialog.

Write down only the changes you kept and why: perhaps quieter keys, one useful EQ cut, gentler lead peaks or a little shared space. If you kept a processor bypassed, record that decision too.

A Tissimo music production teacher can help you compare the dry and processed versions and identify which change actually clarifies the tune.

Free worksheets for this lesson

• Drum Patterns: Sixteen Steps, Four Beats — count the independent kick, snare and hat layers.
• Voice a Chord Loop in MIDI — compare inversions and give the bass a separate register.
• Build a Bassline from Chord Roots — place root-and-fifth attacks against quarter-note kicks.
• Tempo Math: Beats, Bars and Delay Times — calculate eighth, dotted-eighth and other time values.

Sources

• Exporting Audio and Video — Ableton Live 12 manual
• Compare Live editions — Ableton
• Mixing — Ableton Live 12 manual
• Live Audio Effect Reference — Ableton Live 12 manual
• Routing and I/O — Ableton Live 12 manual
• Working with Instruments and Effects — Ableton Live 12 manual
• Audio Fact Sheet — Ableton Live 12 manual
• Basic Gain Staging Overview — Berklee Online
• How high should I set the gain control on my interface? — Focusrite
• BS.1770: programme loudness and true-peak audio level — ITU